Baseline features
Figure 1 shows the flowchart of the article selection. A total of 559 records were identified by the search in Pubmed (n=409), Embase (n=143) and Cochrane (n=20). No restrictions were set on publication status. We could not add any extra articles that were not identified by the literature search after examination of the reference lists. After removal of duplicates (n=92), 467 articles were screened for eligibility based on title and abstract. Of these, 442 records were discarded based on title and abstract because these papers investigated other study outcomes, included animal studies, were reviews or were written in a non-English language that the investigators did not understand. Of the remaining 25 records, full-text articles were retrieved and examined in detail for eligibility. Thirteen additional studies were excluded, because only estimates in figures were presented of spiral artery Doppler indices and no absolute values. We contacted the authors of whom their correct contact information was reported in the articles, to see if they could provide us with more data. Unfortunately, only two authors replied that they could not help us since the research was performed before digitalization and it was not possible to acquire the absolute data of the dated investigations.
A total of 12 studies published in English met the eligibility criteria and were included in this systematic review and meta-analysis. The selected articles were longitudinal prospective (n=7) and cross-sectional studies (n=5). SpA measurements were obtained using transvaginal (n=6), transabdominal ultrasound (n=5) or both (n=1) with probes between 3.5 to 12 MHz. All articles evaluated SpA Doppler during healthy pregnancy. Besides, five articles investigated the differences in SpA Doppler measurements between complicated and healthy pregnancies. There were no articles found that examined SpA Doppler measurements in complicated pregnancies only. Detailed information on the characteristics of included studies is presented in Table 1.
Quality assessment of the included studies
The quality assessment according the modified QUIPS tool per domain is summarized in Table 2. Along these lines, most articles were classified as moderate quality study (n=5) (15, 25-28). Alouini et al. (29), Kurjak et al. (30) and Matijevec et al. (9) were scored low quality, whereas Mäkkikallio, Tekay et al. (31), Hsieh et al. (32) and the two articles of Deurloo et al. (2, 33) scored high quality (Table 2). Ethnicity, non-pregnant BMI and use of medication were poorly described in most studies. In addition, 11 of the 12 included studies failed to report lost to follow up. Intra- and/or inter-observer variation was reported scarcely (n=4).
Meta-analysis
Studies reported Doppler indices at different gestational ages. If one mean SpA Doppler parameter was reported with an interval of gestational weeks as time of measurement, we included the mean of the interval as time of measurement in the meta-analysis (e.g. measurement performed between 6-12 weeks; mean gestational age at measurement 9 weeks). Doppler indices were combined based on measurements during first trimester (0-14 weeks), second trimester (15-27 weeks) or third trimester (28-40 weeks).
Pulsatility index
Ten studies explored the PI at certain (between 5 and 39.5 weeks of gestation) time-points during healthy gestation. The weighted mean of PI decreased from 0.80 (95% CI: 0.70-0.89) (n=9) in the first trimester to 0.50 (95% CI: 0.45-0.55, p<0.001) (n=5) in the second trimester and to 0.49 (95% CI: 0.44-0.53, P=0.752) (n=2) in the third trimester (Table 3, Figure 2 + 2A). 1309 SpA were measured during the first trimester in 778 women. Five studies performed assessment during the second trimester, capturing 1011 SpA measurements in 456 women (2, 9, 30, 32, 33). Two studies published 593 SpA Doppler PI measurements during the third trimester measured in 209 women (30, 32).
Resistance index
Eight studies explored the RI at different time-points during healthy pregnancy (between 6 and 39.5 weeks of gestation). The weighted mean of RI showed a decrease from 0.50 (95% CI: 0.47-0.54) (n=7) in the first trimester to 0.39 (95% CI: 0.37-0.42, p<0.001) (n=4) in the second trimester and 0.36 (95% CI: 0.35-0.38, p=0.037) (n=2) in the third trimester (Table 3, Figure 3 + 3A). Only four studies performed measurements during the second (26, 30-32) and two studies during the third trimester (30, 32). 1076 SpA were measured during the first trimester in 659 women. A total of 820 SpA were measured in 359 women during second trimester and 593 SpA were measured in 209 women during third trimester.
Peak systolic velocity
PSV during the first trimester of healthy pregnancy was explored in four individual studies (between 5 and 39.5 weeks of gestation). The weighted mean of PSV increased from 0.22 m/s (95% CI: 0.13-0.30) (n=3) in the first trimester to 0.28 m/s (95% CI: 0.217-0.40, p=0.373) (n=2) in the second trimester and remained 0.25 m/s (95% CI: 0.20-0.30, p=0.560) (n=2) in the third trimester (Table 3, Figure 4 + 4A). During the first trimester, 614 SpA were measured in 274 women. Hsieh et al. (32) together with Kurjak et al. (30) performed measurements in the second and third trimester, including a total of 209 women and 606, respectively 593 SpA measurements.
Complicated pregnancies
Four articles (2, 15, 26, 28) investigated SpA PI during first trimester in complicated pregnancies (including PE, preterm labour, FGR, pregnancy induced hypertension, anembryonic pregnancies, missed abortion, miscarriage and placental abruption). However, absolute mean PI measurements were only described in two articles during the first trimester (15, 28). Non-significant higher SpA PI measurements were found by Ozkan et al. in 25 women with subsequent miscarriages compared to 189 women having continuing subsequent pregnancies (2.0 (95% CI:1.6-2.7) vs 1.8 (95% CI:1.1-2.9), p=0.320). Results of Ozkaya et al. showed mean SpA PI in 16 pregnancies with adverse outcome (3 miscarriage, 6 missed abortion, 2 preterm labour, 3 IUGR, 1 PE, 1 placental abruption) was 0.97 ± 0.51 (mean ± SD) vs 0.82 ± 0.39 in 84 women with normal pregnancy outcome. The difference in PI between both groups was non-significant. When pooling these data, a statistically significant difference between complicated and healthy pregnancies in the first trimester was found (1.49 vs 0.80, p<0.001, Figure 5 vs Figure 2A). No data was available on PI measurements in second or third trimester pregnancies.
Both articles, along with a third article (27), investigated SpA RI during the first trimester in a total of 51 women. Non-significant differences in RI were found by Ozkan et al. between women with subsequent miscarriages compared to women having continuing subsequent pregnancies (0.7 (95% CI:0.6-0.8) vs 0.7 (95% CI:0.6-0.8), p=0.698). Likewise, Ozkaya et al., found a non-significant difference in SpA RI between women with adverse pregnancy outcome compared to women with normal outcome (0.60 ± 0.37 vs 0.54 ± 0.10). Makikallio et al. described SpA RI during 6, 8, 9 and 11 weeks of gestation in 10 women having PE or preterm labor and compared them with 31 control pregnancies (week 6: 0.82 ± 0.06 vs 0.83 ± 0.06, week 8: 0.84 ± 0.04 vs 0.86 ± 0.06, week 9: 0.85 ± 0.04 vs 0.83 ± 0.06, week 11: 0.75 ± 0.09 vs 0.73 ± 0.09). During the whole study period, no significant differences between both groups were observed. Equally, our pooled meta-analysis data showed no significant differences in RI between complicated and healthy gestation during the first trimester (p=0.568, data not shown). No data was available on RI measurements in second or third trimester pregnancies.
Spiral arteries located in the central part of the placental bed, versus those located at its periphery
Differences between central and peripheral SpA Doppler indices during the second trimester were investigated in two studies (9, 32). Significantly higher PI and RI were found by Matijevic et al. in the SpA located at the peripheral parts of the placenta compared to the central SpA (peripheral SpA PI: 0.48 ± 0.28 (mean ± SD) vs central SpA 0.38 ± 0.26, p<0.001; peripheral SpA RI: 0.38 ± 0.21 vs central SpA 0.32 ± 0.18, p<0.001). No significant differences were found in the PSV between the central and peripheral spiral arteries (p>0.05). Studies of Hsieh et al. concluded that PI and RI values of central placenta bed SpA seemed to be lower, although this finding was statistically non-significant due to the small number of subjects in their sample.
Pooling these results, weighted mean PI in central SpA was 0.46 (95% CI: 0.31-0.62), compared to 0.53 (95% CI: 0.53-0.62) in peripheral SpA (p=0.349) (Figure 6). The weighted mean RI in central SpA was 0.36 (95% CI: 0.29-0.43) and 0.40 (95% CI: 0.39-0.41) in peripheral SpA (p=0.584) (Figure 7).